If the goal is to have a tracking system that follows price tightly, an SMA or EMA with a sufficiently shortened time constant (lookback length) will do just fine. Just remember, whether it is SMA, EMA, Lin Reg, or even Jurik's adaptive EMA, the goal is one in the same. To filter price noise while establishing the trend of the move. The balancing act of noise rejection and delay will always exist, there is no way around it. In addition, even though the Lin Reg is a sort of predictor as you say, the prediction error can be large, causing overshoot and undershoot as it tracks price. The reason is simple: price is not subject to the laws of physics in terms of mechanical momentum andd accceleration. Price can move drastically in the blink of an eye or consolidate in a tight range for seemingly an eternity. Linear predictors such as Kalman and Linear Reg do not apply very well to market action because price is not bandwidth-limited like other engineering systems are. In my earlier days, I've experimented with all types of price-tracking indicators, but ultimately, the SMA and EMA have been all I need to identify the direction of the move. RoughTrader
Hi Cazza. I had a bit of time to do some backtesting. Tweaked the numbers and also added different Linreg periods for Long and Short Entries/Exits, resulting in a profitable system. Equity curve a bit too erratic for me but your system does work, congrats. Appreciate you sharing. --jstox
Another interesting indicator should be the ADX Take a look how it works fine with the Dax future, quite similar to the ER2
I plugged this into Tradestation. It kicks butt on 5 minute ER2 until you add commissions (even with IB's rates). However, after adding Exit on close (so no overnight holding) it does significantly better. Still a very low profit factor however. What's the deal with MACD -- you will only buy if there's a crossover on the same bar?
Who needs indicators? Indicators are for suckers. Indicators fool people. If indicators really worked then we would all be rich because of them! Pick a price, any price. Prices ending in 00, 25, 50, 75 usually work best on any instrument - PSYCHOLOGY. Draw a horizontal line. If price is about to cross above, go long. If price is about to cross below, go short. HINT: Don't go long on a red candle and don't go short on a green candle. Sometimes, you take a small quick profit and sometimes you rake it in. Trading is simple. Indicators take you one or more steps away from the only thing that really matters... PRICE! After you stare at the chart long enough, you'll understand what I am talking about.
Hi guy's, WTF let's share, this is my current favorite using 2 indicators. Backtest this pullback system on 3 minute ER2. The edge on this system is alternating between Longs and Shorts and also not taking pullback entries at extremes. Trades times are Pacific, commission and 1 tick slippage included. --jstox